Uni-Logo
You are here: Home Teaching Summer 2015 Team Project
Document Actions

Team Project: Social Media Analysis

Organizers:

 

Prof. Dr. Peter Fischer

Io Taxidou

Organisational Matters

 

Introductory Meeting: Wednesday April 22th 2015
Room: Building. 51, SR 01-029
Time: 11 am (intoductory meeting, after that scheduled meetings)
Language: German / English

Group Building/Topic Meeting:

Wednesday, April 29th

 

Room: Building. 51, SR 01-029
Time: 11 am

 

Master of Science: 3rd Semester (Teamproject / Masterproject)
ECTS: 16

Overall Topic

 

Social media has begun to dominate many aspects of our everyday lives from information to entertainment up to elaborate remote social relations. People can create content, share it and propagate it at incredible rates, over virtually unlimited distances with no noticeable delay. Furthermore, social media provides new ways to cultivate social relationships in new, often surprising ways, as the rise of services like Facebook (for relationships), Foursquare (for geographic communities) or Instagram (for photos communities) has shown.
 

This enormous wealth of information opens up many new opportunities for both the research community and practical applications. Researches can gain significant knowledge on how humans interact and information is spreading, while individual users as well as service providers may benefit from new, interesting services.
 

In this lab, we develop and implement responsive und scalable methods to analyse media like Twitter. This year's focus will be on classifying user roles in information diffusion

 Individual Topics

  1. Determination of static user roles:

    1. Classification of users along given roles/features
    2. Clustering of users (without known roles)
  2. Streaming clustering of user roles, where they should identify appropriate streaming algorithms that assign users on the fly and recalculate new users interactions.
  3. Evolving user roles: Tracing the evolution of users with identified roles over time
  4. Impact of complex features (structural, users) on cascade lifetime prediction

     

  5. Content-based recommender system using linked data and structured tags
  6. Implementation of optimizations of RDF constraint checking
« July 2017 »
July
MoTuWeThFrSaSu
12
3456789
10111213141516
17181920212223
24252627282930
31
Personal tools